Algorithm for blind identification of MQAM signals in Rayleigh fading channels

An algorithm for blind identifying MQAM (M-ary quadrture amplitude modulation) signals in Rayleigh fading channels was presented to classify MQAM signals easily. The method could classify MQAM signals using the feature parameter which were extracted from two-order, four-order and six-order cumulants without the prior information about received signals. Simulation results show that the classification rate reaches over 90% when SNR (signal-to-noise ratio) is 0 dB, and the algorithm has higher classification rate and better anti-noise ability than other methods. The robustness of the classifier to phase jitter and frequency offset is clearly apparent.